TRAINING METHOD, IMAGE PROCESSING METHOD, DEVICE AND STORAGE MEDIUM FOR GENERATIVE ADVERSARIAL NETWORK

Title:

TRAINING METHOD, IMAGE PROCESSING METHOD, DEVICE AND STORAGE MEDIUM FOR GENERATIVE ADVERSARIAL NETWORK

Document Type and Number:

WIPO Patent Application WO/2020/063648

Kind Code:

A1

Abstract:

A training method, image processing method, computer device and computer-readable storage medium for a generative adversarial network. The training method comprises a generative network training step. The generative network training step comprises: extracting a first resolution sample image from a second resolution sample image (S1); separately providing a first input image and a second input image to a generative network so as to respectively generate a first input image-based first output image and a second input image-based second output image (S2); separately providing the first output image and the second resolution sample image to a discrimination network, and the discrimination network outputting a first output image-based first discrimination result and a second resolution sample image-based second discrimination result (S3); and adjusting parameters of the generative network so as to reduce the loss function of the generative network (S4), thereby obtaining the required images.

Inventors:

LIU HANWEN (CN)
ZHU DAN (CN)
NAVARRETE MICHELINI PABLO (CN)

Application Number:

PCT/CN2019/107761

Publication Date:

April 02, 2020

Filing Date:

September 25, 2019

Export Citation:

Click for automatic bibliography generation Help

Assignee:

BOE TECHNOLOGY GROUP CO LTD (CN)

International Classes:

G06T3/40

Domestic Patent References:

WO2018086354A1

2018-05-17

Foreign References:

CN109345456A	2019-02-15
CN109360151A	2019-02-19
CN109345455A	2019-02-15
CN109255390A	2019-01-22
CN107767343A	2018-03-06
US8675999B1	2014-03-18

Other References: